Functional magnetic resonance imaging (fMRI) has revolutionized the study of the working human brain by providing a sensitive, non-invasive tool for mapping brain activity. The method exploits the sensitivity of the MR signal to local changes in the deoxy-hemoglobin content, called the Blood Oxygenation Level Dependent (BOLD) effect. Yet despite the widespread use of fMRI, the physiological basis of the method--the coupling of neural activity to blood flow and energy metabolismnis still poorly understood. The central physiological phenomenon underlying the BOLD effect is that cerebral blood flow (CBF) increases much more than the cerebral metabolic rate of oxygen (CMRO2), so that local capillary and venous blood are more oxygenated during increased brain activity. We have developed a model to explain this phenomenon (the Oxygen Limitation Model) in which a large change in CBF is required to support a small change in CMRO2. The proposed model is based on two ideas: 1) the function served by a large CBF increase is to maintain mitochondrial pO2 at a constant level, so that oxygen availability does not limit CMRO2, and 2) the mechanism that produces a CBF change with increased neural activity is a feed-forward process that operates without feedback on the current availability of oxygen. The model makes two predictions: 1) the CBF/CMRO2 coupling is reasonably uniform across the brain, and 2) the CBF change with activation deltaF is independent of the baseline CBF. We will test the model in human studies using MRI techniques to measure both CBF and CMRO2. To test the uniformity of CBF/CMRO2 coupling, we will selectively stimulate either the cytochrome oxidase-rich blobs in primary visual cortex or the inter-blob regions, and compare the CBF/CMRO2 coupling curves with each other and with the coupling curve measured in the somatosensory cortex in a finger-tapping experiment. We will test whether deltaF with activation is constant by altering baseline CBF with CO2 and caffeine. Finally we will develop an integrated mathematical model for neurovascular coupling that links the ideas proposed here with several other proposed mechanisms of CBF control.